Sandstrom et al., 24 Applied Optics 472, 1985, describe use of an optical substrate of silicon with a layer of silicon monoxide and a layer of silicon dioxide formed as dielectric films. They indicate that a change in film thickness changes the properties of the optical substrate to produce different colors related to the thickness of the film. That is, the thickness of the film is related to the color observed and a film provided on top of an optical substrate may produce a visible color change. They indicate that a mathematical model can be used to quantitate the color change, and that "[c]alculations performed using the computer model show that very little can be gained in optical performance from using a multilayer structure . . . but a biolayer on the surface changes the reflection of such structures very little since the optical properties are determined mainly by the interfaces inside the multilayer structure . . . . The conclusion is, somewhat surprisingly, that the most sensitive system for detection of biolayers is a single layer coating, while in most other applications performance can be improved by additional dielectric layers."
Sandstrom et al. go on to indicate that slides formed from metal oxides on metal have certain drawbacks, and that the presence of metal ions can also be harmful in many biochemical applications. They indicate that the ideal top dielectric film is a 2-3 nm thickness of silicon dioxide which is formed spontaneously when silicon monoxide layer is deposited in ambient atmosphere, and that a 70-95 nm layer of silicon dioxide on a 40-60 nm layer of silicon monoxide can be used on a glass or plastic substrate. They also describe formation of a wedge of silicon monoxide by selective etching of the silicon monoxide, treatment of the silicon dioxide surface with dichlorodimethylsilane, and application of a biolayer of antigen and antibody. From this wedge construction they were able to determine film thickness with an ellipsometer, and note that the "maximum contrast was found in the region about 65 nm where the interference color changed from purple to blue." They indicate that the sensitivity of such a system is high enough for the detection of protein antigen by immobilized antibodies. They conclude "the designs given are sensitive enough for a wide range of applications. The materials, i.e., glass, silicon, and silicon oxides, are chemically inert and do not affect the biochemical reaction studied. Using the computations above it is possible to design slides that are optimized for different applications. The slides can be manufactured and their quality ensured by industrial methods, and two designs are now commercially available. It is our hope that these sensitive, versatile, and inexpensive tools will further the development of simplified methods in immunology and biochemistry." [Citation omitted.]
Nygren et al., 59 J. Immunol. Methods 145, 1983, describe a system similar to that described above, in which specific anti-human serum albumin (HSA) antibodies are used to detect HSA. FIG. 2 of this publication indicates that 10.sup.-5 mg/ml of HSA can be detected with a 16 hour incubation, but that 10.sup.-6 mg/ml could not be detected in this system. They also state "[a]fter 72 hour of incubation time, the detection limit was lower (down to 1 ng/ml), however, the reaction was then more sensitive to unspecific reactions, giving rise to eventually occurring positive controls."
Nygren et al, U.S. Pat. No. 4,558,012 describe a similar system except that the overall array of layers is adapted to reduce reflection in respect to non-monochromatic or white light in the wavelength range of 525-600 nM.